Code Execution, Shared Snippets, and List Emailing

June 29, 2026

Micro now has code execution for calculations and generative apps, shared snippets and team-shareable context, list emailing, and a smarter LinkedIn inbox.

Code Execution And Generative Apps Are In Beta

Micro can now run code. That means it can do real calculations, transform data, and create working outputs instead of only writing instructions for you to follow.

The bigger version of that is generative apps. Ask Micro to make a dashboard, a deck, an internal tool, or a small app, and it can create an interface you can inspect, preview, and keep iterating on.

These apps can also use the Micro SDK. That means you can build different interfaces on top of your Micro data: a custom view for a pipeline, a reporting dashboard, a lightweight operating tool, or a focused workspace for a specific team workflow.

This is still a beta surface, but the loop is now much more useful: ask for the thing, review it, try it against your data, and share it when it is ready.

Shared Snippets, Context Docs, and Skills

Snippets are now live for repeated writing. Save the language you use often, then insert it from the composer instead of rewriting the same note, intro, follow-up, or answer from scratch.

Snippets are shareable now too. A snippet can become the agreed version of a common response, not just something one person keeps in their own notes.

Context Docs and Skills are also shareable, making it easier to give a team the same reusable background and workflows. You can share the context Micro should know and the ways you want it to work, instead of asking every person to recreate the setup themselves.

List Emailing

Emailing a list is now faster. The composer can pick recipients from a list, so you can write to a saved group without rebuilding the same recipient set each time.

This is especially useful for repeat sends: investor updates, customer follow-ups, recruiting loops, and any workflow where the audience matters as much as the message.

LinkedIn Threads Got Smarter

LinkedIn launched as a first-class surface earlier this month. This update adds more intelligence on top of it.

LinkedIn conversations now carry more useful context in the list itself. Micro can show Status and Labels chips, plus a short summary preview, so you can triage a conversation before opening it.

Those chips are more explainable too. The LinkedIn list can show the supporting context behind an autofilled value, making it easier to understand why a thread was marked as needing a reply, connected to a label, or tied to a workflow.

More Improvements

• Home's Messages section now focuses on messages that need a response and keeps the "View more" path available
• Suggested automations now show the newest suggestions first
• LinkedIn rows handle missing subjects more cleanly and better match the email list layout
• LinkedIn message cards in Home now line up more closely with email
• Subtasks load more reliably when opened from a parent task
• Gmail compose surfaces can match saved signatures to the right sending account more reliably
• Follow-up suggestions now sit inline in chat instead of covering the answer
• Longer chat histories load better from the sidebar and Command-K
• Saved preferences are carried into chat more reliably

Bug Fixes

• AI chat is less likely to duplicate the first user message, lose the current conversation, or let a background stream take over a new chat
• Longer chat lists are less likely to over-fetch or jump while loading older chats
• Thread-bound email drafts now open with the saved body prefilled instead of dropping you into an empty thread
• LinkedIn connect and reconnect states recover more cleanly while a connection is in progress
• App preview resume and shared preview links recover more reliably
• Preview links do a better job opening in the right frame on desktop and avoiding awkward framing on mobile